stemedb/applications/aphoria/src/extractors/option_bounds.rs

use regex::Regex;
use stemedb_core::types::ObjectValue;
use super::{Extractor, build_claim};
use crate::types::{Language, Observation};

/// Detects when Option<T> fields are set to None (unbounded configuration).
///
/// This extractor identifies configuration fields that use Option<T> types
/// and are explicitly set to None in their Default implementation, which
/// often indicates unbounded behavior (e.g., unlimited retries, redirects).
///
/// # Examples
///
/// Detects patterns like:
/// ```rust
/// pub struct Config {
///     pub max_redirects: Option<usize>,  // ← Field declaration
/// }
///
/// impl Default for Config {
///     fn default() -> Self {
///         Self {
///             max_redirects: None,  // ← None assignment (unbounded!)
///         }
///     }
/// }
/// ```
///
/// Creates observation:
/// ```
/// concept_path: "httpclient/max_redirects"
/// predicate: "configured"
/// value: false  // Not configured (allows unbounded)
/// ```
pub struct OptionBoundsExtractor {
    /// Matches: pub field_name: Option<Type>
    field_pattern: Regex,
    /// Matches: field_name: None
    none_pattern: Regex,
}

impl OptionBoundsExtractor {
    /// Create a new OptionBoundsExtractor.
    #[allow(clippy::expect_used)]
    pub fn new() -> Self {
        Self {
            field_pattern: Regex::new(r"pub\s+(\w+):\s*Option<(?:usize|u32|u64|i32|i64|Duration)>")
                .expect("valid regex"),
            none_pattern: Regex::new(r"(\w+):\s*None")
                .expect("valid regex"),
        }
    }

    fn extract_field_names(&self, content: &str) -> Vec<String> {
        self.field_pattern
            .captures_iter(content)
            .map(|cap| cap[1].to_string())
            .collect()
    }

    fn find_none_assignments(&self, content: &str) -> Vec<(String, usize)> {
        content.lines()
            .enumerate()
            .filter_map(|(idx, line)| {
                self.none_pattern.captures(line).map(|cap| {
                    (cap[1].to_string(), idx + 1)
                })
            })
            .collect()
    }
}

impl Default for OptionBoundsExtractor {
    fn default() -> Self {
        Self::new()
    }
}

impl Extractor for OptionBoundsExtractor {
    fn name(&self) -> &str {
        "option_bounds"
    }

    fn languages(&self) -> &[Language] {
        &[Language::Rust]
    }

    fn extract(
        &self,
        path_segments: &[String],
        content: &str,
        _language: Language,
        file: &str,
    ) -> Vec<Observation> {
        let mut observations = Vec::new();

        // Find all Option<T> fields in struct declarations
        let option_fields = self.extract_field_names(content);

        // Find all None assignments in Default impl
        let none_assignments = self.find_none_assignments(content);

        // Match field names: if an Option<T> field is set to None, it's unbounded
        for (field_name, line_num) in none_assignments {
            if option_fields.contains(&field_name) {
                // This is an Option<T> field set to None - unbounded!
                observations.push(build_claim(
                    path_segments,
                    &[&field_name],
                    "configured",
                    ObjectValue::Boolean(false),  // Not configured (unbounded)
                    file,
                    line_num,
                    &format!("{}: None", field_name),
                    0.95,  // High confidence
                    &format!("{} is unbounded (allows None)", field_name),
                ));
            }
        }

        observations
    }

    fn screening_patterns(&self) -> Vec<&str> {
        vec!["Option<", "None"]  // Only run if file has Option types and None
    }

    fn verifiable_predicates(&self) -> Vec<(&str, &str)> {
        vec![
            ("max_redirects", "configured"),
            ("max_retries", "configured"),
            ("max_connections", "configured"),
            ("max_lifetime", "configured"),
            ("idle_timeout", "configured"),
            ("pool_size", "configured"),
        ]
    }
}

#[cfg(test)]
mod tests {
    use super::*;

    #[test]
    fn test_detects_none_assignment() {
        let content = r#"
            pub struct Config {
                pub max_redirects: Option<usize>,
            }

            impl Default for Config {
                fn default() -> Self {
                    Self {
                        max_redirects: None,
                    }
                }
            }
        "#;

        let extractor = OptionBoundsExtractor::new();
        let obs = extractor.extract(
            &["httpclient".to_string(), "config".to_string()],
            content,
            Language::Rust,
            "config.rs",
        );

        assert_eq!(obs.len(), 1);
        assert_eq!(obs[0].predicate, "configured");
        assert_eq!(obs[0].value, ObjectValue::Boolean(false));
        assert!(obs[0].concept_path.contains("max_redirects"));
    }

    #[test]
    fn test_detects_multiple_none_assignments() {
        let content = r#"
            pub struct Config {
                pub max_redirects: Option<usize>,
                pub max_retries: Option<u32>,
            }

            impl Default for Config {
                fn default() -> Self {
                    Self {
                        max_redirects: None,
                        max_retries: None,
                    }
                }
            }
        "#;

        let extractor = OptionBoundsExtractor::new();
        let obs = extractor.extract(&[], content, Language::Rust, "config.rs");

        assert_eq!(obs.len(), 2);
        assert!(obs.iter().any(|o| o.concept_path.contains("max_redirects")));
        assert!(obs.iter().any(|o| o.concept_path.contains("max_retries")));
    }

    #[test]
    fn test_ignores_non_option_fields() {
        let content = r#"
            pub struct Config {
                pub timeout: u64,
            }

            impl Default for Config {
                fn default() -> Self {
                    Self {
                        timeout: 30,
                    }
                }
            }
        "#;

        let extractor = OptionBoundsExtractor::new();
        let obs = extractor.extract(&[], content, Language::Rust, "config.rs");
        assert_eq!(obs.len(), 0);  // Should not detect non-Option fields
    }

    #[test]
    fn test_ignores_some_assignments() {
        let content = r#"
            pub struct Config {
                pub max_redirects: Option<usize>,
            }

            impl Default for Config {
                fn default() -> Self {
                    Self {
                        max_redirects: Some(10),
                    }
                }
            }
        "#;

        let extractor = OptionBoundsExtractor::new();
        let obs = extractor.extract(&[], content, Language::Rust, "config.rs");
        assert_eq!(obs.len(), 0);  // Should not detect Some(_) assignments
    }

    #[test]
    fn test_screening_patterns() {
        let extractor = OptionBoundsExtractor::new();
        let patterns = extractor.screening_patterns();
        assert!(patterns.contains(&"Option<"));
        assert!(patterns.contains(&"None"));
    }

    #[test]
    fn test_verifiable_predicates() {
        let extractor = OptionBoundsExtractor::new();
        let predicates = extractor.verifiable_predicates();
        assert!(predicates.contains(&("max_redirects", "configured")));
        assert!(predicates.contains(&("max_retries", "configured")));
    }
}